COMMENTARY

Trimorphic stepping stones pave the way to fungal virulence

Robert J. Bastidas and Joseph Heitman1 Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, NC 27710

X he fungal kingdom encompasses Pseudohyphal growth Ϸ1.5 million species (1) as di- A verse as single-celled yeasts, pathogens of animals/plants, and Tplant root symbionts. Fungi are eukary- Y Hyphal growth otic, closely aligned with metazoans (2, 3). Animals and fungi diverged Ϸ1 billion years ago; their last common ancestor was unicellular, motile, and aquatic. Some fungi grow as unicellular yeasts, but most B X X are filamentous multicellular organisms. Importantly, some fungi are dimorphic, growing as both yeast and filamentous forms (i.e., Saccharomyces cerevisiae). Yet Fig. 1. Fungal morphological transitions. (A) Yeast undergoes dimorphic transitions to pseudohyphae or others are trimorphic and grow as yeast, hyphae controlled by different elements (X, Y). (B) A continuous transition from yeast to pseudohyphae hyphae, and pseudohyphae (i.e., Candida to hyphae controlled by a dosage-dependent factor (x, X). albicans). S. cerevisiae pseudohyphal growth long eluded detection—it requires special conditions/strains and was lost dur- yeasts (lacking Cph1 and Efg1 transcrip- filaments were observed. When the tet- tion factors or cyclin Hgc1) or filaments UME6 strain was grown as hyphae (doxϪ) ing domestication (4). How pseudohyphae ϩ are related to yeast and hyphae (as a dis- (lacking the Tup1 repressor) are both and then shifted to repress (dox ), filaments tinct fate or a continuum) was unknown avirulent, linking both forms to pathogen- produced a majority of pseudohyphae by 3 h until the report of Carlisle et al. (5)inthis esis (8–10). Subsequently, a strain was and yeast by 7 h. Thus both filamentous issue of PNAS. They reveal that engineered in which morphogenesis is modes can be reversibly evoked by express- pseudohyphae are intermediate between controlled by regulated expression of a ing a single regulatory element in a dosage- yeast and hyphae, with implications for filamentation repressor, Nrg1, with the tet dependent fashion, providing evidence that pathogen–host interactions and fungal promoter (11). Cells grown without doxy- pseudohyphae are intermediate between evolution. cycline express Nrg1 and grow as yeast, yeast and hyphae, rather than a distinct fate. C. albicans is a trimorphic fungus from whereas growth with doxycycline re- When animals were infected with yeast of Ϫ the Ascomycota, diverged from S. cerevisiae pressed Nrg1 and filamentous growth en- the tet-UME6 strain (dox ), increased Ume6 Ϸ200 million years ago (6, 7), and a com- sued. Animals infected with yeast re- expression led to enhanced hyphal growth mensal of the human microbiota of the gas- mained healthy yet harbored a significant and tissue invasion and more rapid demise; trointestinal tract, mucous membranes, and latent fungal burden in the kidney. Add- animals given doxycycline survived much skin. C. albicans infects humans, causing oro- ing doxycycline to drinking water acti- longer. Hence, hyphal growth (or hyphal- pharyngeal disease, vaginitis, and systemic vated filamentation, with progression to specific gene expression) promotes C. albi- life-threatening infection. Serum induces C. lethal infection. These studies provide ro- cans virulence. albicans yeast to produce germ tubes, and bust support for the concept that dimor- The findings of Carlisle et al. (5) have growth in defined media elicits two filamen- phic transitions underlie C. albicans viru- broad implications for dimorphism in fungal tous growth modes: pseudohyphae and hy- lence, and they show that yeast can pathogenesis (Fig. 2). S. cerevisiae and C. phae (Fig. 1A). Pseudohyphae resemble hy- penetrate tissues, whereas hyphae are nec- albicans both undergo yeast–pseudohyphal phae, but are morphologically distinct. In essary for progression to lethal infection. transitions, but only C. albicans develops both modes the normally ovoid yeast cells In cultured macrophages S. cerevisiae is hyphae. As C. albicans evolved into a suc- are elongated, but in pseudohyphae each killed after phagocytosis, whereas C. albi- cessful commensal of the mammalian gastro- cell–cell junction is constricted, and the di- cans yeast switch to hyphae, killing and intestinal tract, formation of hyphae in bio- ameter between cell walls is wider in the escaping macrophages (12). films was likely necessary to compete with middle than the ends. In contrast, hyphal cell The studies of Carlisle et al. (5) involve bacteria, or to form cooperative multispecies walls are parallel with no mother–daughter controlling expression of Ume6, a zinc-finger biofilms (15). Hyphae are also critical for C. or septal junction constrictions. Pseudohy- transcription factor downstream of Nrg1 in albicans to survive and escape host macro- phae and hyphae also differ with respect to circuits for dimorphic transition and viru- phages. As a result, the trimorphic species C. septin localization and nuclear division (7). lence (13, 14). In their studies, Ume6 was albicans is a successful commensal and Given these differences, it has been unclear expressed from the doxycycline-regulatable pathogen, whereas the dimorphic yeast S. whether pseudohyphae are an intermediate promoter (tet-UME6), leading to controlled between yeast and hyphae or are an alterna- consequences: yeast with the promoter off tive fate (Fig. 1). This question, and implica- (doxϩ), and hyphae in response to high-level Author contributions: R.J.B. and J.H. wrote the paper. tions for virulence, were addressed by Carl- Ume6 (doxϪ), even under non-filament- The authors declare no conflict of interest. isle et al. with engineered strains (5). inducing conditions. Intermediate expression See companion article on page 599. The C. albicans dimorphic yeast–hy- elicited a third fate: pseudohyphae. Increas- 1To whom correspondence should be addressed. E-mail: phae transition is thought to underlie its ing Ume6 levels converted pseudohyphae to [email protected]. success as a pathogen. Mutants locked as hyphae, and hybrid hyphal/pseudohyphal © 2009 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0811994106 PNAS ͉ January 13, 2009 ͉ vol. 106 ͉ no. 2 ͉ 351–352 Downloaded by guest on September 24, 2021 envision an ancestral fungus as a unicellular yeast making the leap, first to pseudohyphae and then hyphae. This would have facilitated foraging for mating partners and nutrients and rapid colonization of habitats, including plants/animals. In this model mutations may block just hyphal growth, or hyphal and pseudohyphal growth. Overexpression of other regulatory components may enable dimorphic yeasts to produce hyphae. Most fungi are filamentous with no yeast form, thus mutations preventing the dimorphic return to yeast likely confer selective bene- fits. These studies may provide insights into how closely related species evolved divergent life styles (Fig. 2). S. cerevisiae grows as yeast and pseudohyphae but not hyphae, whereas its close relative Ashbya gossypii grows only as hyphae and Holleya sinecauda grows as yeast, pseudohyphae, and hyphae (22–24). A. gossypii thus lacks functions for hyphal–yeast Fig. 2. Evolution of fungal form and function. Mono-, di-, and trimorphic fungi are shown. transition, and H. sinecauda has factors en- abling hyphal growth from yeast and pseudohyphae. Recent studies reveal that the cerevisiae is an uncommon pathogen (16). iomycota phylum, the plant pathogen Usti- pescadillo ortholog promotes production of Exploring genetic circuits controlling dimor- lago maydis can grow as yeast, but successful budding yeast cells from C. albicans hyphae phism in both species, and those between plant infection requires filamentous dikary- (25), and related factors may enable Crypto- them, is likely to be illustrative. ons produced by mating (20). The human coccus neoformans hyphae to produce yeast The ability of fungi to infect humans oc- pathogen Cryptococcus neoformans under- cells (21). Other fungi, such as Mucor sp. in curred independently multiple times, and goes a similar sexual dimorphic transition, the Zygomycota, prominently grow as fila- successful human pathogens are diverse (Fig. but spores or dried yeast cells infect humans mentous fungi in lab aerobic conditions yet 2). The prominent role for hyphae in C. albi- and only yeast occur in the host (21). Other also grow as multibudded yeasts in anaero- cans virulence is contrasted with a leading fungi infecting plants and humans are strictly bic/high-CO2 conditions (26). Conditions or role for yeast in other pathogens. The di- filamentous. How these diverse pathogenic mutations enabling other filamentous fungi morphic fungal pathogens (i.e., Histoplasma strategies subvert host defenses may share to grow as pseudohyphae or yeast may re- capsulatum) are thermally dimorphic, grow- common principles, or converge via distinct main to be discovered. Understanding ge- ing as filamentous molds at lower environ- mechanisms. netic circuits evoking transitions in fungal mental temperatures and as yeast at 37 °C The findings of Carlisle et al. (5) have form, and impact on biology and infection, (17). Humans are infected by inhaled broader evolutionary implications. The find- will continue to fascinate for years to come. conidia that must convert to yeast to be ing that pseudohyphae are a developmental ACKNOWLEDGMENTS. This work was supported pathogenic. Strains or mutants locked as way station between yeast and hyphae sug- by National Institutes of Health Grants AI39115 filaments are avirulent (18, 19). In the Basid- gests stepwise evolution (Fig. 1B). We can and AI50113.

1. James TY, et al. (2006) Reconstructing the early evolu- 9. Zheng X, et al. (2004) Hgc1, a novel hypha-specific G1 18. Maresca B, et al. (1977) Regulation of dimorphism in tion of fungi using a six-gene phylogeny. Nature cyclin-related protein regulates Candida albicans hy- the pathogenic fungus Histoplasma capsulatum. Na- 443:818–822. phal morphogenesis. EMBO J 23:1845–1856. ture 266:447–448. 2. Baldouf SL, Palmer JD (1993) Animals and fungi are 10. Braun BR, et al. (2000) Identification and characteriza- 19. Nemecek JC, et al. (2006) Global control of dimorphism each other’s closest relatives: Congruent evidence from tion of TUP1-regulated genes in Candida albicans. Ge- and virulence in fungi. Science 312:583–588. multiple proteins. Proc Natl Acad Sci USA 90:11558– netics 156:31–44. 20. Kamper J, et al. (2006) Insights from the genome of the 11562. 11. Saville SP, et al. (2003) Engineered control of cell mor- biotrophic fungal plant pathogen Ustilago maydis. Na- 3. Wainright PO, et al. (1993) Monophyletic origins of the phology in vivo reveals distinct roles for yeast and ture 444:97–101. metazoa: An evolutionary link with fungi. Science filamentous forms of Candida albicans during infec- 21. Hull CM, Heitman J (2002) Genetics of Cryptococcus 260:340–342. tion. Eukaryotic Cell 2:1053–1060. neoformans. Annu Rev Genet 36:557–615. 4. Gimeno CJ, et al. (1992) Unipolar cell divisions in the 12. Lorenz MC, Fink GR (2002) Life and death in a macro- 22. Wendland J, Walther A (2005) Ashbya gossypii: a model phage: role of the glyoxylate cycle in virulence. Eu- yeast Saccharomyces cerevisiae lead to filamentous for fungal developmental biology. Nat Rev Microbiol karyot Cell 1:657–662. growth: regulation by starvation and RAS. Cell 3:421–429. 13. Kadosh D, Johnson AD (2005) Induction of the Candida 68:1077–1090. 23. Dietrich FS, et al. (2004) The Ashbya gossypii genome as albicans filamentous growth program by relief of tran- 5. Carlisle PL, et al. (2009) Expression levels of a fila- a tool for mapping the ancient Saccharomyces cerevi- scriptional repression: a genome-wide analysis. Mol ment-specific transcriptional regulator are sufficient siae genome. Science 304:304–307. Biol Cell 16:2903–2912. to determine Candida albicans morphology and vir- 14. Banerjee M, et al. (2008) UME6, a novel filament- 24. Gastmann S, et al. (2007) A molecular toolbox for ulence. Proc Natl Acad Sci USA 106:599–604. specific regulator of Candida albicans hyphal extension manipulating Eremothecium coryli. Microbiol Res 6. Berman J, Sudbery PE (2002) Candida albicans:amo- and virulence. Mol Biol Cell 19:1354–1365. 162:299–307. lecular revolution built on lessons from budding yeast. 15. Nobile CJ, Mitchell AP (2007) Microbial biofilms: e plu- 25. Shen J, Cowen LE, Griffin AM, Chan L, Ko¨hler JR (2008) Nat Rev Genet 3:918–930. ribus unum. Curr Biol 17:R349–353. The Candida albicans pescadillo homolog is required for 7. Sudbery P, et al. (2004) The distinct morphogenic 16. McCusker JH (2006) in Molecular Principles of Fungal normal hypha-to-yeast morphogenesis and yeast prolif- states of Candida albicans. Trends Microbiol 12:317– Pathogenesis, eds Heitman J, Filler SG, Edwards JE, Mitch- eration. Proc Natl Acad Sci USA 105:20918–20923. 324. ell AP (Am Soc Microbiol, Washington, DC), pp 245–259. 26. Bartnicki-Garcia S, Nickerson WJ (1962) Induction of 8. Lo HJ, et al. (1997) Nonfilamentous C. albicans mutants 17. Klein BS, Tebbets B (2007) Dimorphism and virulence in yeast-like development in Mucor by carbon dioxide. J are avirulent. Cell 90:939–949. fungi. Curr Opin Microbiol 10:314–319. Bacteriol 84:829–840.

352 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0811994106 Bastidas and Heitman Downloaded by guest on September 24, 2021